Sensitivity of a Riparian Large Woody Debris Recruitment Model to the Number of Contributing Banks and Tree Fall Pattern

Riparian large woody debris (LWD) recruitment simulations have traditionally applied a random angle of tree fall from two well-forested stream banks. We used a riparian LWD recruitment model (CWD, version 1.4) to test the validity these assumptions. Both the number of contributing forest banks and predominant tree fal l d irect ion s igni f icant ly inf luenced s imulated r iparian LWD del ivery , but there was no apparent interact ion between these factors. Pooled across al l treatments , the average predicted 300-year cumulative LWD recruitment was 77.1 m”/lOO m reach with both banks forested compared to 49.3 m’/lOO m reach when only one s ide was t imbered. Total recrui tment within bank cover categories (one versus both forested) depended on the directionality of the falling stem. When only one bank was forested, the CWD model predicted the same riparian LWD recruitment for the random and CWD default tree fall patterns (-39 m3/100 m reach), the pattern biased toward the channel yielded twice this volume, a pattern quartering toward the channel produced 64% more LWD, and the pattern paralleling the channel contributed almost 30% less than random. With both banks forested, the random, default, and quartering simulations resulted in similar delivery (about 78 m3/100 m reach), the pattern biased toward the channel contributed almost 14% more LWD, and the parallel pattern yielded 26% less. Because CWD is similar in design and operation to other riparian LWD recruitment models, it follows that any simulation of wood delivery to streams should be checked for their consis tency wi th local forest cover and tree fai lure pat terns . West. J. Appl. For. 19(2):117-122.